Immune memory is critical for pathogen resistance as well as for vaccine efficacy; it is a defining feature of adaptive immunity. To deepen our fundamental understanding of how immune memory works we have focused on elucidating the inherent differences between memory B cells (MBC) and nave B cells (NBC). A key limitation in studying murine MBC is their vanishingly small numbers and the difficulty in identifying and purifying them; hence, studies on function and gene expression have been quite limited. To overcome this barrier our lab has developed multiple systems by which to obtain and purify large numbers of homogeneous, bona fide MBC, allowing us to assay them in vivo and in vitro. In the last period we used this unique capability to de- fine subsets of murine MBC, revealed by expression of CD80, PD-L2, and CD73. We focused on the three subsets defined by CD80 and PD-L2: double positive (DP), PD-L2-single positive (SP), and double negative (DN). Strikingly, upon reimmunization DP MBC are potent in generating AFCs but cannot generate GCs, whereas DN MBC make GC but lag at making AFCs; SPs are intermediate. This heterogeneity is present with- in IgM MBC or IgG MBC compartments and hence subset identity rather than isotype controls function. This finding was exciting because it explained how MBC could provide rapid effector function while still reseeding the memory compartment for future responses. It also revealed that studying unseparated mixtures of MBC could yield misleading conclusions. Our long-term goal is to understand the mechanisms that underlie this different behavior among the MBC subsets and also between MBC and NBC overall. We will address this at multiple levels, taking both genetic and functional approaches. As a first step, we have recently obtained microarray-based data on mRNA expression of the MBC subsets. This is a good start in terms of defining the reprogramming that underlies differential functions. This, along with our recently published functional data, form the basis for our proposal. However, we have at best an incomplete database of both intrinsic gene expression differences and no information on epigenetic differences. Similarly, we have as yet only a rudimentary idea of functional differences. We will address both of these issues in Aims 1 and 2 respectively. In Aim 1 we will investigate via a combination of RNA- and ChIP-seq of resting MBC subsets, as well as by RNA-seq of stimulated MBC, the intrinsic differences among MBC subsets and NBC. In Aim 2 we will delineate functional differences among these same cells both in vitro and in vivo. We expect thereby to add functional meaning to the genetic studies and reciprocally to understand the different phenotypes we may see by virtue of the genetic/epigenetic data. Our recently obtained microarray data identified transcription factors that are differentially expressed. In Aim 3 we will delete two of these genes, zbtb32 and klf2, in established MBC, using a new inducible Cre mouse, and then test the functional consequences. Overall this work will be both hypothesis generating and hypothesis testing and will provide important insights into immune memory.